Package: weka
Version: 3.4.4
Revision: 1
Source: http://schwehr.org/software/weka/%n-%v.tar.bz2
Source2: mirror:sourceforge:weka/Tutorial.pdf
Source3: mirror:sourceforge:weka/ExplorerGuide.pdf
Source4: http://www.cs.waikato.ac.nz/~ml/weka/weka.bn.pdf
Source5: mirror:sourceforge:weka/Experiments.pdf
Source-MD5: c1598f9e60a45152e0e698a75a2cd372
Source2-MD5: 787a145b668449b75f4e6f38d13ad06e
Source3-MD5: 8b5011a68e9a8e84a9bd3d3b7787926f
Source4-MD5: 63156c0c2de3da34be46b3414965f70e
Source5-MD5: f1555562d177c9b27b592a79c015afa8
Description: Machine learning/pattern recognition in java
Depends: system-java14
BuildDepends: ant, system-java14-dev
CompileScript: <<
  (cd weka && ant)
  (cd weka && ant docs)
  (cd weka && ant exejar)
  (cd weka && ant remotejar)
  find . -name CVS | xargs rm -r
<<
InstallScript: <<
	# Make a quick executable for the GUI
	mkdir -p %i/bin
	echo '#!/bin/bash' > %i/bin/weka
	echo "java -jar %p/share/java/weka/weka.jar" >> %i/bin/weka
	chmod +x %i/bin/weka

	mkdir -p %i/share/doc/%n
	cp -r doc %i/share/doc/%n/doc
	cp -r ../ExplorerGuide.pdf ../Tutorial.pdf ../weka.bn.pdf ../Experiments.pdf %i/share/doc/%n/
	cp -r wekadocs/data %i/share/doc/%n
<<
SplitOff: <<
	Package: %N-doc
	Description: Docs on machine learning and pattern recognition
	Depends: %N
	Files: share/doc/%N/*.pdf share/doc/%N/doc
<<
JarFiles: dist/weka.jar dist/remoteExperimentServer.jar
DescDetail: <<
Weka is a collection of machine learning algorithms for data mining
tasks. The algorithms can either be applied directly to a dataset or
called from your own Java code. Weka contains tools for data
pre-processing, classification, regression, clustering, association
rules, and visualization. It is also well-suited for developing new
machine learning schemes.

To start the GUI, just run 'weka'

For a quick command line sample run, try this:

	java weka.core.Instances /sw/share/doc/weka/data/weather.arff
	java weka.classifiers.trees.J48 -t /sw/share/doc/weka/data/weather.arff

<<
DescPort: <<
Do NOT use Source: mirror:sourceforge:weka/weka-%n.zip
This is a prebuilt set of jars which is NOT ok with the fink rules

The distribution does not come with the ant build instructions.
Talked to the authors and agreed to do a cvs tagged pull of the tree:

export CVSROOT=:pserver:cvs_anon@cvs.scms.waikato.ac.nz:/usr/local/global-cvs/ml_cvs
cvs login
# empty pass
cvs co -r stable-3-4-4 weka
cvs co -r stable-3-4-4 tests
cvs co -r stable-3-4-4 wekadocs

This does have the build.xml for ant

The data directory has never been in the cvs tree, so I just made
another tar for it.

This info file is under the GPL and/or what ever other license the
fink team deams best. -Kurt
<<
License: GPL
Maintainer: Kurt Schwehr <goatbar@users.sourceforge.net>
Homepage: http://www.cs.waikato.ac.nz/~ml/weka/